Nwanosike Victor

I’m a detail-oriented Data Analyst and Document Annotation Specialist with 2+ years of experience validating complex datasets and ensuring data integrity across high-volume, multi-field reporting pipelines. I specialize in field-by-field document comparisons—matching extracted values back to source PDFs, structured JSON exports, and tabular financial data.

In my current role, I validate data outputs against source documents to confirm accuracy across hundreds of data points per report. I review structured data using predefined JSON schemas, verify nested value placement, and check mathematical and logical correctness of line-item totals and cross-references before finalization.

I’ve also contributed directly to improving AI extraction quality by flagging edge cases, ambiguous cases, and recurring extraction challenges. I document and escalate issues to senior consultants, then help drive iterative process improvements that reduce annotation errors over time.

Previously, I built ETL workflows to profile, cleanse, and restructure legacy datasets using SQL and Power Query, improving downstream accuracy by ~40% and reducing errors. I also authored internal data governance policies for quality standards, classification, and compliance, strengthening consistency across business units.